ワンクリックで
architecture
TRIGGER when presenting Architectural options, Architecting systems, or discussing tradeoffs. Append to any Prompt.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
メニュー
TRIGGER when presenting Architectural options, Architecting systems, or discussing tradeoffs. Append to any Prompt.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes
Broad-to-narrow research Orchestration - start with a wide census, harden the scope from returned Evidence, and specialize Subagents only once the exact situation is proven. TRIGGER when the Architect says "progressive research", "start broad", or when dispatching research or debugging Subagents into a situation whose cause or shape is unknown. DO NOT TRIGGER when the exact scope is already proven - dispatch the specialized Subagent per /subagents directly.
Consolidate the knowledge wiki — revisit journaled topics, fold repeated findings into each topic's Claude.md, surface contradictions, keep cross-links true. TRIGGER on a scheduled pass or when a topic tree has grown messy or redundant. DO NOT TRIGGER to file fresh session material (that is journal), or to gather new research.
File a session's loose material into the knowledge wiki — each capture into its owning topic folder, the topic's Claude.md kept current. TRIGGER after a session hands you loose captures (notes, clippings, findings, screenshots) that need a home in the wiki. DO NOT TRIGGER to gather new research, or to consolidate the wiki itself (that is dream).
Operate Dent via first-party API. Trigger: Dent API, Funnels, Offers, Pages, Articles, Courses, Spaces, Analytics. Do not trigger for WordPress-only work.
Autonomous async Execution. The Architect hands off and walks away; the Agent drives the Task to done — makes every Architectural call via /trace + /pcc + ranking, executes through /subagents, verifies from each implementing Subagent's Evidence (they exercise their own flows, real browser for UI), and returns finished code plus the recorded Decisions for review. TRIGGER when the Architect says "code-first", "/code-first", "execute with /subagents", "execute directly & autonomously", "drive it and report back", "iterate until done", "I'm going to bed", "off to bed", or signals an AFK / overnight handoff. DO NOT TRIGGER when the Architect wants options before action — that is the default mode (propose via /pcc and wait).
| name | architecture |
| description | TRIGGER when presenting Architectural options, Architecting systems, or discussing tradeoffs. Append to any Prompt. |
State the problem, what triggered it, and the highest Architectural Decision before lower-level choices.
Defaults, naming, edge cases, and implementation details wait until the data model, module boundary, public API, and dependency direction are approved. Never: ask about defaults before the data model, naming before the Architecture, edge cases before the happy path, or implementation details before the approach.
Architecture is module boundaries, public contracts, data ownership, dependency direction, new modules, and schema mutations. Convention follows repo Precedent. Implementation is the Agent's tactical work: method internals, error messages, and control flow. Never: present an implementation choice as an Architectural Decision.
Build toward the Architecture the code should have. Current code is evidence, not a wall around the Decision.
Name the shape you think is right in one sentence, plus the Plan and why. Show real names, public API methods, the call site, or pseudocode where useful. Never: describe the Architecture only as an abstract category.
Trace the code it touches and find the case it cannot handle, the boundary that does not hold, the capability the User would lose, or the caller it forces you to rewrite. When using Subagents, ask them to break the proposed Architecture and propose fixes. Never: ask a Subagent to summarize the current code when the Task is to harden a proposed Architecture.
A weak spot is a defect in the Architecture, not a tradeoff to list and move past. Fold in fixes and attack again until the shape is coherent, elegant, and functional. Never: unbroken hypothesis, status-quo wall, or noted-not-fixed.
Use /show-architecture to show what exists and what changes.
The option must show what the caller writes to use it, not only the data model. Example: show the public method call and payload shape beside the ownership change. Never: only diagram tables while hiding the public API the caller will use.
Names are Architecture when they enter files, public APIs, database schema, or reusable vocabulary. Never: invent identifiers inside an Architectural Example.
Every option needs pros, cons, and confidence so the Architect can compare it.
Each option must solve at least one problem the others do not. State what each option is best for; if two options are best for the same thing, merge them. Never: two options that are the same idea.
Good Architecture eliminates future Decisions. Name what choosing the option forces the project to decide next. Example: "Choosing this makes the entity own the public API; controllers stop assembling the payload."
An option must be viable in real Execution. Do not present exits from the direction the Architect is exploring. Never: "defer", "use an external service" for a simple thing, "code-only" when runtime control is needed, "keep current", "start over", "abandon this direction", or any option you would not recommend.
Ask what each system knows, what it does not know, and who owns each concept.
Do not couple unrelated concerns. Same shape does not mean same concern when lifecycles differ. Example: billing is not tenancy; plans are not feature flags. Never: make one system change whenever an unrelated system changes.
Compare lines of code and maintenance burden against revenue potential at 1,000 Users: retention, upsells, and reduced churn.